[SCSI] sd_dif: problem with verify of type 1 protection information (PI)
authorJeremy Higdon <jeremy@sgi.com>
Thu, 18 Apr 2013 09:55:23 +0000 (02:55 -0700)
committerJames Bottomley <JBottomley@Parallels.com>
Thu, 2 May 2013 22:37:03 +0000 (15:37 -0700)
It appears to me that there is a problem with handling of type 1 protection
information.

It is considering a logical block reference tag of 0xffffffff to be an error,
but it is actually valid any time ((lba & 0xffffffff) == 0xffffffff) [for
example, 2TiB-1, 4TiB-1, 6TiB-1, etc.].

I'm going by what's written in 4.18.3 of SBC3, where there doesn't appear
to be any invalid value for the reference tag.

Signed-off-by: Jeremy Higdon <jeremy@sgi.com>
Acked-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
drivers/scsi/sd_dif.c

index 04998f36e5071bdda94ff225f39b0ca3616eeb7b..6174ca4ea27594487d7dc0828d9e21841742b8ed 100644 (file)
@@ -93,14 +93,6 @@ static int sd_dif_type1_verify(struct blk_integrity_exchg *bix, csum_fn *fn)
                if (sdt->app_tag == 0xffff)
                        return 0;
 
-               /* Bad ref tag received from disk */
-               if (sdt->ref_tag == 0xffffffff) {
-                       printk(KERN_ERR
-                              "%s: bad phys ref tag on sector %lu\n",
-                              bix->disk_name, (unsigned long)sector);
-                       return -EIO;
-               }
-
                if (be32_to_cpu(sdt->ref_tag) != (sector & 0xffffffff)) {
                        printk(KERN_ERR
                               "%s: ref tag error on sector %lu (rcvd %u)\n",